KR100195118B1 - Image enhancing method based on the quantized mean-separate histogram equalization having the gain control function and circuit thereof - Google Patents

Abstract

The image quality improving method and the circuit according to the present invention quantize the level of the input video signal, average the input video signal by screen unit, and quantize the average obtained to output the quantized average level, and output the quantized video signal. The image is divided into a predetermined number of quantized sub-images according to the quantized average level, the cumulative density function value is obtained for each quantized sub-image based on the gray level distribution, and the cumulative density function obtained for the input image signal and the quantized sub-image. Output the interpolated cumulative density function value for each sub-image by interpolation based on the value, output the improved signal by independently histogram equalization for each sub-image based on the interpolated cumulative density function value for each sub-image, and input Depending on the amount of gray level change between the video signal and the improved signal and the level of the input video signal While improving the contrast hameuroseo regulate hwaryang kept constant on the average brightness of a given image, and also by preventing an artifact of the excessive improvement improve image quality.

Description

Image Quality Improvement Method Using Quantized Average-Separated Histogram Equalization with Gain Control and Its Circuit

The present invention relates to a method for improving image quality using a quantized average-separated histogram equalization having a gain control function and a circuit thereof. In particular, the present invention relates to quantizing an input video signal and dividing the quantized video signal into a predetermined number of sub-images. The present invention relates to a method and a circuit for improving image quality by adjusting histogram equalization independently of an image to improve contrast while preventing an artifact caused by excessive improvement.

The basic operation of histogram equalization is to convert a given input image based on the histogram of the input image, where a histogram represents a gray level distribution in a given input image.

This gray level histogram provides an overall depiction of the appearance of the image. Gray levels appropriately adjusted according to the sample distribution of the image improve the appearance or contrast of the image.

Among many methods for improving contrast, histogram equalization, the method of improving the contrast of a given image according to the sample distribution of the image, is the most widely known and described in [1] and [2]: [1] JSLim , Two-Dimensional Signal and Image Processing, Prentice Hall, Englewood Cliffs, New Jersey, 1990, [2] RC Gonzalez and P. Wints, Digital Image Processing, Addison-Wesley, Reading, Massachusetts, 1977.

In addition, useful applications of histogram equalization methods, including medical image processing and radar image processing, are disclosed in [3] and [4] below: [3] J. Zimmerman, S. Pizer, E. Staab, E. Perry, W. McCartney, and B. Brenton, Evaluation of the effectiveness of adaptive histogram equalization for contrast enhancement, IEEE Tr.on Medical Imaging, pp. 304-312, Dec. 1988, [4] Y.Li, W. Wang , and DYYu, Application of adaptive histogram equalization to x-ray chest image, Proc. of the SPIE, pp. 513-514, vol. 2321,1994.

Therefore, a technique using a histogram of a given image is usefully applied in various fields such as medical image processing, infrared image processing, and radar image processing.

However, a problem that may occur in contrast improvement based on histogram equalization for a given image may include a large difference between input and output. In other words, the contrast can be excessively improved by histogram equalization based on an algorithm that depends on the histogram of the input image, which causes a problem of providing an image of an undesired form.

In addition, the conventional histogram equalization circuit requires a configuration capable of storing all occurrences of all gray levels, resulting in a problem of high hardware cost. For example, assuming that gray level L is L = 256, 256 memory elements are required to store the occurrence counts of all levels, and 256 accumulators are required to accumulate the occurrence counts of all levels.

An object of the present invention is to quantize an input image, divide the quantized input image into a predetermined number of sub-images, and independently histogram equalize the divided sub-images to improve contrast while maintaining the average brightness constant. The present invention provides a method of improving image quality by adjusting a gain of an adaptively improved signal according to a signal level.

Another object of the present invention is to quantize an input image, divide the quantized input image into a predetermined number of sub-images, and independently histogram equalize the divided sub-images to improve contrast while maintaining the average brightness constant. The present invention provides a circuit that improves image quality by adjusting a gain of an adaptively improved signal according to a level of an image signal.

In order to achieve the above object, the image quality improving method according to the present invention is a method for improving image quality by histogram equalization of a video signal represented by a predetermined number of gray levels: (a) quantizing the level of the input video signal step; (b) outputting a quantized average level by quantizing the average obtained by obtaining an average of an input video signal in units of screens; (c) dividing the quantized video signal into a predetermined number of quantized sub-pictures according to the quantized average level in step (a); obtaining a cumulative density function value for each of the quantized sub-images based on the gray level distribution; (e) outputting an interpolated cumulative density function value for each sub-image by interpolation based on an input video signal and a cumulative density function value obtained for each of the quantized sub-images; (f) outputting an improved signal by independently histogram equalization for each subimage based on the cumulative density function value interpolated for each subimage; And (g) adjusting the change amount according to the gray level change amount between the input video signal and the improved signal and the level of the input video signal.

In order to achieve the above another object, an image quality improvement circuit according to the present invention is a circuit for improving image quality by histogram equalizing a video signal represented by a predetermined number of gray levels: quantizing the level of an input video signal First quantization means for outputting a video signal; First calculating means for calculating a gray level distribution of the quantized video signal in units of screens; Second calculating means for calculating an average level of the input video signal in units of screens; Second quantization means for quantizing the average level and outputting a quantized average level; Third calculating means for dividing the gray level distribution calculated by the first calculating means into a predetermined number of quantized sub-images according to the quantized average level and calculating a cumulative density function value for each quantized sub-image; Interpolation means for outputting an interpolated cumulative density function value for each sub-image by interpolation based on an input video signal and a cumulative density function value calculated for each of the quantized sub-images; Output means for mapping the input image signal to a gray level according to a cumulative density function value calculated for each of the quantized sub-images and outputting an improved signal; And gain control means for controlling the gain of the improved signal according to the amount of change of the gray level between the input video signal and the improved signal and the level of the input video signal.

1 illustrates an example of quantizing an L-level discrete signal into a Q-level discrete signal in order to explain the quantization concept of the present invention.

2 is a view for explaining an interpolation concept applied to the present invention.

3 is a block diagram according to an embodiment of an image quality improvement circuit using quantized average-separated histogram equalization with gain control according to the present invention.

4 is a block diagram according to another embodiment of an image quality improvement circuit using quantized average-separated histogram equalization with gain control according to the present invention.

A method of improving image quality using quantized mean-separate histogram equalization with a gain control function proposed by the present invention will be described.

First, the quantized mean-separated histogram equalization algorithm will be described.

The given image {X} consists of L discrete gray levels {X0, X1, ..., XL-1}, where X0 = 0 is the black level and XL-1 = 1 the white level. Indicates.

The original discrete input levels {X0, X1, ..., XL-1} are quantized to Q discrete levels defined as {Z0, Z1, ..., ZQ-1}, where ZQ-1 = XL- Assume that 1 is Q 1 L and that {Z0, Z1, ..., ZQ-1} 1C {X0, X1, ..., XL-1}.

Thus, an example of quantizing the L-level discrete signal to the Q-level discrete signal is shown in FIG.

Q [Xk] is called a quantization operation and is defined as follows.

Q [Xk] = Zq if Zq-1 XK 1 Zq

When {Z} = Q [{X}] and Zm = Q [Xm], Xm represents the average level of the original image, {Z} represents the quantized input image, Zm represents the quantized average level, The quantized input image {Z} is divided into two sub-images {Z} L and {Z} U around Zm. Here, all samples in the quantized subimage {Z} L are less than or equal to the quantized average level Zm, and all samples in the quantized subimage {Z} U are larger than the quantized average level Zm.

Each of the quantized probability density functions PDF of the sub-images {Z} L and {Z} U may be represented by Equations 1 and 2 below.

[Equation 1]

[Equation 2]

Where PL (Zq) is the probability of the q th quantized gray level Zq in the quantized sub-image {Z} L, and PU (Zq) is the q th quantized gray level Zq in the quantized sub-image {Z} U. Where NqL and NqU represent the number of quantized sub-pictures {Z} L and quantized sub-pictures {Z} U, and the number of levels (Zq) appears in the quantized sub-pictures {Z} U, respectively. , The total number of samples of each quantized sub-image {Z} U.

At this time, the cumulative density function (CDF) of each of the quantized sub-image {Z} L and the quantized sub-image {Z} U is defined as in Equations 3 and 4 below.

[Equation 3]

[Equation 4]

Where CL (Zm) = 1 and CU (ZQ-1) = 1.

The interpolated cumulative density functions cL (Xk), cU (Xk) can be approximately calculated through linear interpolation from CL (Zq) and CU (Zq).

As shown in FIG. 2, assuming Q [Xk] = Zq 1 Zm, Z-1 = 0, and cL (Xk) is interpolated as in Equation 5 below.

[Equation 5]

Similarly, assuming Q [Xk] = Zq Zm, cU (Xk) is interpolated as in Equation 6 below.

[Equation 6]

Here, CU (Zm) = 0.

Finally, based on the interpolated cumulative density function, the output YH of the proposed quantized mean-separated histogram equalization is given by Equation 7 for the given input image Xk.

[Equation 7]

Where Zm '= Zm + XL-1 / (L-1), which is the next gray level after Zm in {X0, X1, ..., XL-1}.

Next, an algorithm for adjusting the gain of the signal improved by the quantized mean-separated histogram equalization will be described.

The basic concept of gain control is to limit the change of the maximum gray level of the input image signal (Yi = Xk) according to the degree of contrast improvement when the contrast is improved by using quantized average-separated histogram equalization.

First, the relationship between the input image signal Yi = Xk and the improved signal YH can be expressed by Equation 8 or Equation 9 below.

[Equation 8]

or,

[Equation 9]

Is the amount of change (degree of improvement) produced by quantized mean-separated histogram equalization, that is, the level of the input image signal Yi and the level YH mapped to the new gray level by quantized mean-separated histogram equalization. Say the difference.

In order to prevent excessive improvement by histogram equalization, the change amount Δ is limited in the present invention as follows.

[Equation 10]

1 | △ 1 | 1 gf (Yi)

Where f (Yi) is defined as the maximum bounding function, where f (Yi) is a function of input (Yi) and is always positive, that is, f (Yi) 1 0, g (g 1 0) is a gain control parameter.

Equation 10 may be expressed as Equation 11 as an equivalent equation.

[Equation 11]

g-f (Yi) 1 Δ 1 g-f (Yi)

The concept of limiting the amount of change Δ shown in Equation 10 is related to Weber's ratio. In fact, if f (Yi) = Yi, the following Equation 12 can be obtained from Equation 10 above.

[Equation 12]

here, Is the amount corresponding to the webber ratio. This Weber's ratio is an experimental fact meaning that when Y1 changes by gY1 and Y2 changes by gY2, humans feel that the degree of change is the same.

Therefore, the concept of gain adjustment applied to the present invention is to control the gain, i.e., the degree of improvement, of the improved signal by using the quantized mean-separated histogram based on the webber ratio.

Δ 'is defined as a limited amount of change and can be expressed as in Equation 13.

[Equation 13]

That is, if the absolute value of the change amount △ is less than gf (Yi), the change amount △ is used as the limited change amount △ ', and if the change amount △ is larger than gf (Yi), it is limited to gf (Yi). When the change amount Δ is smaller than -gf (Yi), the gain of the improved signal YH is controlled by limiting to -gf (Yi).

Therefore, the final output signal (YOUT) of the present invention can be represented by the following equation (14).

[Equation 14]

3 and 4, an embodiment of an image quality improvement circuit using quantized average-separated histogram equalization with a gain adjustment function will be described.

3 is a block diagram according to an embodiment of an image quality improvement circuit using quantized average-separated histogram equalization with gain control according to the present invention.

In FIG. 3, the first quantizer 102 quantizes the input video signal Xk having the L discrete level to the Q discrete level to output the quantized image Zq. The frame histogram calculator 104 calculates a gray level distribution of the quantized input image Zq in units of one screen. Here, the screen unit may be a field but is a frame.

The frame average calculator 106 calculates an average level Xm of the input video signal in units of one frame. The second quantizer 108 quantizes the average level Xm of the input video signal Xk and outputs the quantized average level Zm.

The divider 110 quantizes a predetermined number (here two) based on the quantized gray level distribution calculated by the frame histogram calculator 104 based on the quantized average level Zm output from the second quantizer 108. The probability density functions PL (Zq) and PU (Zq) of the two quantized sub-images are output by dividing the sub-images {Z} L and {Z} U into two sub-images. PL (Zq) and PU (Zq)) may be calculated by Equations 1 and 2 above.

Here, all samples in the quantized subimage {Z} L are less than or equal to the quantized average level Zm, and all samples in the quantized subimage {Z} U are larger than the quantized average level Zm.

The first CDF calculator 112 inputs the probability density function PL (Zq) of the quantized sub-image {Z} L having all image samples below the quantized average level from the divider 110 to quantize the sub-image {Z}. The cumulative density function value of L is calculated using Equation 3 above.

The second CDF calculator 114 inputs the quantization by inputting the probability density function PU (Zq) of the quantized sub-image {Z} U whose all image samples output from the divider 110 are larger than the quantized average level Zm. The cumulative density function value of the sub-image {Z} U is calculated using Equation 4 above.

The CDF memory 116 stores the cumulative density function values CL (Zq) and CU (Zq) of the quantized sub-images {Z} L, {Z} U calculated by the first and second CDF calculators 112 and 114. Is updated in units of screens according to the synchronization signal SYNC, and the cumulative density function values CL (Zq) and CU (Zq) before one screen stored during the update are supplied to the first and second interpolators 118 and 120. . Here, the synchronization signal SYNC becomes a field synchronization signal when the screen unit is a field, becomes a frame synchronization signal when it is a frame, and the CDF memory 116 is used as a buffer.

The first interpolator 118 inputs the cumulative density function of the quantized sub-image {Z} L and the input image signal XK and linearly interpolates the interpolated cumulative density function value cL (Xk) by Equation 5 above. )

The first mapper 120 inputs the interpolated cumulative density function cL (Xk), the input image signal Xk, and the quantized average level Zm, thereby sub-image {Z} L which is equal to or less than the quantized average level Zm. Samples of are mapped to gray levels from X0 to Zm according to the interpolated cumulative density function cL (Xk).

The second interpolator 122 inputs the cumulative density function CU (Zq) of the quantized sub-image {Z} U and the input image signal Xk and linearly interpolates the interpolated cumulative density function according to Equation 6 above. Outputs (cU (Xk)).

The second mapper 124 inputs the interpolated cumulative density function cU (Xk), the input image signal Xk, and the quantized average level Zm, so that the sub-image {Z} is larger than the quantized average level Zm. Samples of U are mapped to gray levels from Zm 'to XL-1 according to the interpolated cumulative density function cU (Xk). Where Zm '= Zm + XL-1 / (L-1).

That is, the video signal Xk inputted to the first and second interpolators 118 and 122 is the video signal of the next frame of the video signal Xk inputted to the first quantizer 102 and the frame average calculator 104. . However, the hardware can be reduced by omitting the frame memory by using the property of having a high correlation between adjacent frames.

The selector 126 compares the input image signal Xk and the quantized average level Zm output from the second quantizer 108 and compares the first image signal Xk with the first quantized average level Zm. The mapper 120 is selected, otherwise the second mapper 124 is selected to output the improved signal YH, which can be represented by the above equation (7).

Here, the first quantizer 102 to the first selector 126 may be referred to as quantized mean-separated histogram equalizer 100.

Also, without using the frame histogram calculator 104 and the CDF calculators 112 and 114 separately, the gray level distribution of the image signal divided by the CDF calculators 112 and 114 without the frame histogram calculator 102 is calculated and the CDF is calculated based on this. Can be calculated

Meanwhile, the subtractor 202 of the gain controller 200 subtracts the input video signal Yi = Xk from the improved signal YH output from the first selector 126 to quantize the average-separated histogram equalizer 100. The change amount?

The gain characteristic determiner 204 uses the maximum limit function f (Yi), which is a function of the input video signal Yi, to limit the improvement of the input video signal Yi. That is, the marginal improvement amount of the input video signal Yi is calculated.

For example, when f (Yi) = K, where K is a constant, this limits the input improvement equally regardless of the input gray value, the maximum limit function (Where a is a preset constant) or (Where a is a preset constant), which otherwise limits the amount of improvement of the input image according to the input gray level value.

The second selector 206 compares the input image signal Xk with the quantized average level Zm output from the second quantizer 108 and compares the input image signal Xk with the average level Zm. Below, the first gain control parameter gL is selected, otherwise the second gain control parameter gU is selected. Here, the first gain control parameter gL is a parameter for the sub picture signal that is less than or equal to the quantized average level Zm, and the second gain control parameter gU is for a sub picture signal that is greater than the quantized mean level Zm. Parameter.

The multiplier 208 multiplies the maximum limit function f (Yi) value output from the gain characteristic determiner 204 by the gain control parameter g selected by the second selector 206 to thereby limit the limit value gf (Yi). ) Here, the limit g g f (Y i) is a characteristic of the gain controller 200 of the present invention.

The limiter 210 limits the amount of change Δ by comparing the amount of change Δ output from the subtractor 202 with the limit value gf (Yi) output from the multiplier 208, thereby limiting the amount of change Δ (△ ') is output.

The adder 212 outputs the output signal Yout of the gain controller 200 as shown in Equation 14 by adding the input image signal Yi and the limited change amount Δ ′ output from the limiter 210. do.

FIG. 4 is a block diagram according to another embodiment of an image quality improvement circuit using quantized average-separated histogram equalization with a gain control function according to the present invention. A detailed description thereof will be omitted.

Therefore, a description will be given of the gain controller 200 ′ different from that shown in FIG. 3.

In Fig. 4, the subtractor 222 of the gain controller 200 'subtracts the input image signal Yi = Xk from the improved signal YH output from the selector 126 to quantize the average-separated histogram equalizer. The change amount? Produced by (100) is obtained. The gain characteristic determiner 224 limits the improvement of the input image signal Yi by using the maximum limit function f (Yi) which is a function of the input image signal Yi.

The multiplier 226 multiplies the maximum limit function f (Yi) value output from the gain characteristic determiner 224 and the input gain control parameter g to output the limit value gf (Yi).

Here, the gain control parameter g is given the same value with respect to the input video signal, unlike the gain control parameter given for each quantized sub-image in FIG. 3.

The limiter 228 compares the change amount? Output from the subtractor 222 with the limit value gf (Yi) output from the multiplier 226 to limit the change amount? And limit the change amount as shown in Equation 13 above. (△ ') is output.

The adder 230 adds the input video signal Yi and the limited change amount? 'Output from the limiter 228 to output the output signal Yout of the gain controller 200' as shown in Equation 14 above. Output

The present invention can be applied to a wide range of fields related to the improvement of image quality of a video signal. That is, the present invention can be applied to broadcast equipment, radar signal processing systems, medical engineering, home appliances, and the like.

As described above, the present invention provides an improved signal gain based on the ratio of webber and the effect of maintaining the overall brightness of a given image while improving contrast by effectively reducing the sudden brightness change and artifacts that occur in conventional histogram equalization. By adjusting it, there is an effect of improving the image quality by preventing artifacts caused by excessive improvement in contrast during histogram equalization.

In addition, the circuit of the present invention quantizes the input video signal and independently histogram equalizes the sub-picture divided into a predetermined number to store and accumulate only the number of occurrences of the quantized level for CDF calculation, thereby simplifying hardware and reducing cost. It works.

Claims (30)

A method for improving image quality by histogram equalization of video signals represented by a predetermined number of gray levels: (a) quantizing the level of an input video signal; (b) outputting a quantized average level by quantizing the average obtained by obtaining an average of an input video signal in units of screens; (c) dividing the quantized video signal into a predetermined number of quantized sub-pictures according to the quantized average level in step (a); obtaining a cumulative density function value for each of the quantized sub-images based on the gray level distribution; (e) outputting an interpolated cumulative density function value for each sub-image by interpolation based on an input video signal and a cumulative density function value obtained for each of the quantized sub-images; (f) outputting an improved signal by independently histogram equalization for each subimage based on the cumulative density function value interpolated for each subimage; And and adjusting the change amount according to the gray level change amount between the input video signal and the improved signal and the level of the input video signal. The method of claim 1, wherein the step (c) divides the quantized video signal into two sub-pictures according to the quantized average level. The method of claim 1, wherein the interpolation in the step (e) is linear interpolation. The image quality improving method according to claim 1, further comprising the step (e1) of delaying the input video signal in units of screens and outputting the delayed video signal in the step (e). The method of claim 1, wherein step (f) (f1) mapping samples of each quantized subimage to gray levels according to the cumulative density function value interpolated for each subimage; And and selecting one of the signals mapped to the gray level for each of the quantized sub-images according to a result of comparing and comparing the input video signal with the quantized average level. 2. The method of claim 1, wherein the adjusting of the gain of the improved signal in step (g) is based on a ratio of webbers. The method of claim 1, wherein step (g) (g1) subtracting the input video signal from the improved signal to detect a change amount corresponding to a difference; (g2) calculating a limit improvement amount of the input video signal by a predetermined maximum limit function and a predetermined gain control parameter; (g3) outputting a limited change amount by limiting the change amount according to a result of comparing and comparing the limit improvement amount with the change amount; And (g4) adding the limited amount of change to the input video signal. 8. The method of claim 7, wherein, in the step (g2), the improvement of the input video signal is limited by a predetermined maximum limit function. The method of claim 7, wherein step (g2) (g21) outputting a limited video signal by limiting improvement of the input video signal by a predetermined maximum limit function; And (g22) multiplying the limited video signal by a predetermined gain control parameter and outputting a marginal improvement amount. 10. The method of claim 9, wherein the predetermined gain control parameter comprises a plurality of gain control parameters for each quantized sub-image. 10. The method of claim 9, wherein the predetermined gain control parameter is a gain control parameter that is equally applied to an input video signal. The method of claim 9, wherein the maximum limit function f (Y i) is And a is a preset constant. The method of claim 9, wherein the maximum limit function f (Y i) is And a is a preset constant. 10. The method of claim 9, wherein in step (g3), when the absolute value of the change amount is less than or equal to the limit function value, the change amount is output as the limited change amount, and otherwise, the limit function value is output as the limited change amount. How to improve image quality. A circuit for improving image quality by histogram equalization of a video signal represented by a predetermined number of gray levels: First quantization means for quantizing the level of an input video signal and outputting a quantized video signal; First calculating means for calculating a gray level distribution of the quantized video signal in units of screens; Second calculating means for calculating an average level of the input video signal in units of screens; Second quantization means for quantizing the average level and outputting a quantized average level; Third calculating means for dividing the gray level distribution calculated by the first calculating means into a predetermined number of quantized sub-images according to the quantized average level and calculating a cumulative density function value for each quantized sub-image; Interpolation means for outputting an interpolated cumulative density function value for each sub-image by interpolation based on an input video signal and a cumulative density function value calculated for each of the quantized sub-images; Output means for mapping the input image signal to a gray level according to a cumulative density function value calculated for each of the quantized sub-images and outputting an improved signal; And And gain control means for controlling the gain of the improved signal according to the amount of change in gray level between the input video signal and the improved signal and the level of the input video signal. The image quality improvement circuit according to claim 15, wherein the screen unit is a frame unit and the predetermined number is two. 16. The apparatus of claim 15, further comprising a screen memory for delaying the input video signal by a screen unit in order to input the video signal having the same frame and the cumulative density function calculated by the third calculating means to the interpolation means. Image quality improvement circuit. The method of claim 15, further comprising: a buffer for updating the cumulative density function calculated for each sub-image quantized by the third calculating means in units of screens, and supplying the cumulative density function value stored during the updating to the interpolation means. An image quality improvement circuit. The image quality improving circuit of claim 15, wherein the interpolation is linear interpolation. The method of claim 15, wherein the output means A first mapper for mapping the input image signal to a gray level of a first range according to a cumulative density function value of a quantized sub-image corresponding to the first sub-image below a quantized average level; A second mapper for mapping the input image signal to a gray level of a second range according to a cumulative density function value of the quantized sub-image corresponding to the second sub-image larger than the quantized average level; And And a selector for comparing the input video signal with the quantized average level to select a first mapper if the first sub-image is selected, and otherwise select a second mapper. The method of claim 17, wherein the output means A first mapper that maps the gray level signal to a gray level according to a cumulative density function value of the quantized sub-picture corresponding to the first sub-picture below the quantized average level; A second mapper for mapping a gray level of a second range according to a cumulative density function value of a quantized sub-picture corresponding to the second sub-picture larger than the quantized average level; And And a selector for comparing the video signal output from the screen memory with the quantized average level and selecting a first mapper if the first sub image is selected, and otherwise selecting a second mapper. 16. The image quality improvement circuit according to claim 15, wherein the gain control means controls the gain of the improved signal based on the Weberby. 16. The apparatus of claim 15, wherein the gain control means Detection means for subtracting the input video signal from the improved signal to detect a change amount corresponding to a difference; Gain characteristic determining means for calculating a maximum limit function value according to the level of the input video signal to limit the improvement of the input video signal by a predetermined maximum limit function; Selecting means for comparing the input video signal with a quantized average level and selecting a first gain control parameter if the input video signal is equal to or less than a quantized average level; otherwise, selecting means for selecting one of second gain control parameters; Multiplication means for multiplying the maximum limit function value by a gain control parameter selected by the selection means to output a limit function value; Limiting means for outputting a limited change amount by limiting the change amount according to a result of comparing the limit function value with the change amount; And And adding means for adding a limited amount of change to the input video signal to output a signal in which the gain of the improved signal is adjusted. The method of claim 23, wherein the maximum limit function f (Yi) is And a is a preset constant. The method of claim 23, wherein the maximum limit function f (Yi) is And a is a preset constant. 24. The method of claim 23, wherein the limiting means outputs the change amount as the limited change amount as it is, when the absolute value of the change amount is equal to or less than the limit function value, and otherwise outputs the limit function value as the limited change amount. Image quality improvement circuit. 16. The apparatus of claim 15, wherein the gain control means Detection means for subtracting the input video signal from the improved signal to detect a change amount corresponding to a difference; Gain characteristic determining means for calculating a maximum limit function value according to the level of the input video signal to limit the improvement of the video signal input by a predetermined maximum limit function; Multiplication means for multiplying the maximum limit function value by a gain control parameter of a predetermined value to output a limit function value; Limiting means for outputting a limited change amount by limiting the change amount according to a result of comparing the limit function value with the change amount; And And adding means for adding a limited amount of change to the input video signal to output a signal in which the gain of the improved signal is adjusted. The method of claim 27, wherein the maximum limit function f (Yi) is And a is a preset constant. The method of claim 27, wherein the maximum limit function f (Yi) is And a is a preset constant. 28. The method of claim 27, wherein the limiting means outputs the change amount as the limited change amount as it is, when the absolute value of the change amount is equal to or less than the limit function value, and otherwise outputs the limit function value as the limited change amount. Image quality improvement circuit.